Commits · e39f63fe0d94ddc8a54929e90ecf93078538d63b · Kirill Smelkov / linux

27 Apr, 2022 5 commits

net: remove comments that mention obsolete __SLOW_DOWN_IO · e39f63fe

Bjorn Helgaas authored Apr 25, 2022

The only remaining definitions of __SLOW_DOWN_IO (for alpha and ia64) do
nothing, and the only mentions in networking are in comments.  Remove these
mentions.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

e39f63fe

net: wan: atp: remove unused eeprom_delay() · dac173db

Bjorn Helgaas authored Apr 25, 2022

atp.h is included only by atp.c, which does not use eeprom_delay().  Remove
the unused definition.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

dac173db

net: tls: fix async vs NIC crypto offload · c706b2b5

Jakub Kicinski authored Apr 25, 2022

When NIC takes care of crypto (or the record has already
been decrypted) we forget to update darg->async. ->async
is supposed to mean whether record is async capable on
input and whether record has been queued for async crypto
on output.
Reported-by: Gal Pressman <gal@nvidia.com>
Fixes: 3547a1f9 ("tls: rx: use async as an in-out argument")
Tested-by: Gal Pressman <gal@nvidia.com>
Link: https://lore.kernel.org/r/20220425233309.344858-1-kuba@kernel.orgSigned-off-by: Jakub Kicinski <kuba@kernel.org>

c706b2b5

net: dsa: mt753x: fix pcs conversion regression · fae46308

Russell King (Oracle) authored Apr 25, 2022

Daniel Golle reports that the conversion of mt753x to phylink PCS caused
an oops as below.

The problem is with the placement of the PCS initialisation, which
occurs after mt7531_setup() has been called. However, burited in this
function is a call to setup the CPU port, which requires the PCS
structure to be already setup.

Fix this by changing the initialisation order.

Unable to handle kernel NULL pointer dereference at virtual address 0000000000000020
Mem abort info:
  ESR = 0x96000005
  EC = 0x25: DABT (current EL), IL = 32 bits
  SET = 0, FnV = 0
  EA = 0, S1PTW = 0
  FSC = 0x05: level 1 translation fault
Data abort info:
  ISV = 0, ISS = 0x00000005
  CM = 0, WnR = 0
user pgtable: 4k pages, 39-bit VAs, pgdp=0000000046057000
[0000000000000020] pgd=0000000000000000, p4d=0000000000000000, pud=0000000000000000
Internal error: Oops: 96000005 [#1] SMP
Modules linked in:
CPU: 0 PID: 32 Comm: kworker/u4:1 Tainted: G S 5.18.0-rc3-next-20220422+ #0
Hardware name: Bananapi BPI-R64 (DT)
Workqueue: events_unbound deferred_probe_work_func
pstate: 60000005 (nZCv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--)
pc : mt7531_cpu_port_config+0xcc/0x1b0
lr : mt7531_cpu_port_config+0xc0/0x1b0
sp : ffffffc008d5b980
x29: ffffffc008d5b990 x28: ffffff80060562c8 x27: 00000000f805633b
x26: ffffff80001a8880 x25: 00000000000009c4 x24: 0000000000000016
x23: ffffff8005eb6470 x22: 0000000000003600 x21: ffffff8006948080
x20: 0000000000000000 x19: 0000000000000006 x18: 0000000000000000
x17: 0000000000000001 x16: 0000000000000001 x15: 02963607fcee069e
x14: 0000000000000000 x13: 0000000000000030 x12: 0101010101010101
x11: ffffffc037302000 x10: 0000000000000870 x9 : ffffffc008d5b800
x8 : ffffff800028f950 x7 : 0000000000000001 x6 : 00000000662b3000
x5 : 00000000000002f0 x4 : 0000000000000000 x3 : ffffff800028f080
x2 : 0000000000000000 x1 : ffffff800028f080 x0 : 0000000000000000
Call trace:
 mt7531_cpu_port_config+0xcc/0x1b0
 mt753x_cpu_port_enable+0x24/0x1f0
 mt7531_setup+0x49c/0x5c0
 mt753x_setup+0x20/0x31c
 dsa_register_switch+0x8bc/0x1020
 mt7530_probe+0x118/0x200
 mdio_probe+0x30/0x64
 really_probe.part.0+0x98/0x280
 __driver_probe_device+0x94/0x140
 driver_probe_device+0x40/0x114
 __device_attach_driver+0xb0/0x10c
 bus_for_each_drv+0x64/0xa0
 __device_attach+0xa8/0x16c
 device_initial_probe+0x10/0x20
 bus_probe_device+0x94/0x9c
 deferred_probe_work_func+0x80/0xb4
 process_one_work+0x200/0x3a0
 worker_thread+0x260/0x4c0
 kthread+0xd4/0xe0
 ret_from_fork+0x10/0x20
Code: 9409e911 937b7e60 8b0002a0 f9405800 (f9401005)
---[ end trace 0000000000000000 ]---
Reported-by: Daniel Golle <daniel@makrotopia.org>
Tested-by: Daniel Golle <daniel@makrotopia.org>
Fixes: cbd1f243 ("net: dsa: mt7530: partially convert to phylink_pcs")
Signed-off-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Link: https://lore.kernel.org/r/E1nj6FW-007WZB-5Y@rmk-PC.armlinux.org.ukSigned-off-by: Jakub Kicinski <kuba@kernel.org>

fae46308

net: generalize skb freeing deferral to per-cpu lists · 68822bdf

Eric Dumazet authored Apr 22, 2022

Logic added in commit f35f8219 ("tcp: defer skb freeing after socket
lock is released") helped bulk TCP flows to move the cost of skbs
frees outside of critical section where socket lock was held.

But for RPC traffic, or hosts with RFS enabled, the solution is far from
being ideal.

For RPC traffic, recvmsg() has to return to user space right after
skb payload has been consumed, meaning that BH handler has no chance
to pick the skb before recvmsg() thread. This issue is more visible
with BIG TCP, as more RPC fit one skb.

For RFS, even if BH handler picks the skbs, they are still picked
from the cpu on which user thread is running.

Ideally, it is better to free the skbs (and associated page frags)
on the cpu that originally allocated them.

This patch removes the per socket anchor (sk->defer_list) and
instead uses a per-cpu list, which will hold more skbs per round.

This new per-cpu list is drained at the end of net_action_rx(),
after incoming packets have been processed, to lower latencies.

In normal conditions, skbs are added to the per-cpu list with
no further action. In the (unlikely) cases where the cpu does not
run net_action_rx() handler fast enough, we use an IPI to raise
NET_RX_SOFTIRQ on the remote cpu.

Also, we do not bother draining the per-cpu list from dev_cpu_dead()
This is because skbs in this list have no requirement on how fast
they should be freed.

Note that we can add in the future a small per-cpu cache
if we see any contention on sd->defer_lock.

Tested on a pair of hosts with 100Gbit NIC, RFS enabled,
and /proc/sys/net/ipv4/tcp_rmem[2] tuned to 16MB to work around
page recycling strategy used by NIC driver (its page pool capacity
being too small compared to number of skbs/pages held in sockets
receive queues)

Note that this tuning was only done to demonstrate worse
conditions for skb freeing for this particular test.
These conditions can happen in more general production workload.

10 runs of one TCP_STREAM flow

Before:
Average throughput: 49685 Mbit.

Kernel profiles on cpu running user thread recvmsg() show high cost for
skb freeing related functions (*)

    57.81%  [kernel]       [k] copy_user_enhanced_fast_string
(*) 12.87%  [kernel]       [k] skb_release_data
(*)  4.25%  [kernel]       [k] __free_one_page
(*)  3.57%  [kernel]       [k] __list_del_entry_valid
     1.85%  [kernel]       [k] __netif_receive_skb_core
     1.60%  [kernel]       [k] __skb_datagram_iter
(*)  1.59%  [kernel]       [k] free_unref_page_commit
(*)  1.16%  [kernel]       [k] __slab_free
     1.16%  [kernel]       [k] _copy_to_iter
(*)  1.01%  [kernel]       [k] kfree
(*)  0.88%  [kernel]       [k] free_unref_page
     0.57%  [kernel]       [k] ip6_rcv_core
     0.55%  [kernel]       [k] ip6t_do_table
     0.54%  [kernel]       [k] flush_smp_call_function_queue
(*)  0.54%  [kernel]       [k] free_pcppages_bulk
     0.51%  [kernel]       [k] llist_reverse_order
     0.38%  [kernel]       [k] process_backlog
(*)  0.38%  [kernel]       [k] free_pcp_prepare
     0.37%  [kernel]       [k] tcp_recvmsg_locked
(*)  0.37%  [kernel]       [k] __list_add_valid
     0.34%  [kernel]       [k] sock_rfree
     0.34%  [kernel]       [k] _raw_spin_lock_irq
(*)  0.33%  [kernel]       [k] __page_cache_release
     0.33%  [kernel]       [k] tcp_v6_rcv
(*)  0.33%  [kernel]       [k] __put_page
(*)  0.29%  [kernel]       [k] __mod_zone_page_state
     0.27%  [kernel]       [k] _raw_spin_lock

After patch:
Average throughput: 73076 Mbit.

Kernel profiles on cpu running user thread recvmsg() looks better:

    81.35%  [kernel]       [k] copy_user_enhanced_fast_string
     1.95%  [kernel]       [k] _copy_to_iter
     1.95%  [kernel]       [k] __skb_datagram_iter
     1.27%  [kernel]       [k] __netif_receive_skb_core
     1.03%  [kernel]       [k] ip6t_do_table
     0.60%  [kernel]       [k] sock_rfree
     0.50%  [kernel]       [k] tcp_v6_rcv
     0.47%  [kernel]       [k] ip6_rcv_core
     0.45%  [kernel]       [k] read_tsc
     0.44%  [kernel]       [k] _raw_spin_lock_irqsave
     0.37%  [kernel]       [k] _raw_spin_lock
     0.37%  [kernel]       [k] native_irq_return_iret
     0.33%  [kernel]       [k] __inet6_lookup_established
     0.31%  [kernel]       [k] ip6_protocol_deliver_rcu
     0.29%  [kernel]       [k] tcp_rcv_established
     0.29%  [kernel]       [k] llist_reverse_order

v2: kdoc issue (kernel bots)
    do not defer if (alloc_cpu == smp_processor_id()) (Paolo)
    replace the sk_buff_head with a single-linked list (Jakub)
    add a READ_ONCE()/WRITE_ONCE() for the lockless read of sd->defer_list
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>
Link: https://lore.kernel.org/r/20220422201237.416238-1-eric.dumazet@gmail.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>

68822bdf

26 Apr, 2022 4 commits

net: usb: qmi_wwan: add support for Sierra Wireless EM7590 · 56121548

Ethan Yang authored Apr 25, 2022

add support for Sierra Wireless EM7590 0xc081 composition.
Signed-off-by: Ethan Yang <etyang@sierrawireless.com>
Acked-by: Bjørn Mork <bjorn@mork.no>
Link: https://lore.kernel.org/r/20220425054028.5444-1-etyang@sierrawireless.comSigned-off-by: Paolo Abeni <pabeni@redhat.com>

56121548

net/af_packet: add VLAN support for AF_PACKET SOCK_RAW GSO · dfed913e

Hangbin Liu authored Apr 25, 2022

Currently, the kernel drops GSO VLAN tagged packet if it's created with
socket(AF_PACKET, SOCK_RAW, 0) plus virtio_net_hdr.

The reason is AF_PACKET doesn't adjust the skb network header if there is
a VLAN tag. Then after virtio_net_hdr_set_proto() called, the skb->protocol
will be set to ETH_P_IP/IPv6. And in later inet/ipv6_gso_segment() the skb
is dropped as network header position is invalid.

Let's handle VLAN packets by adjusting network header position in
packet_parse_headers(). The adjustment is safe and does not affect the
later xmit as tap device also did that.

In packet_snd(), packet_parse_headers() need to be moved before calling
virtio_net_hdr_set_proto(), so we can set correct skb->protocol and
network header first.

There is no need to update tpacket_snd() as it calls packet_parse_headers()
in tpacket_fill_skb(), which is already before calling virtio_net_hdr_*
functions.

skb->no_fcs setting is also moved upper to make all skb settings together
and keep consistency with function packet_sendmsg_spkt().
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Acked-by: Willem de Bruijn <willemb@google.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Link: https://lore.kernel.org/r/20220425014502.985464-1-liuhangbin@gmail.comSigned-off-by: Paolo Abeni <pabeni@redhat.com>

dfed913e

net: dsa: ksz: added the generic port_stp_state_set function · de6dd626

Arun Ramadoss authored Apr 24, 2022

The ksz8795 and ksz9477 uses the same algorithm for the
port_stp_state_set function except the register address is different. So
moved the algorithm to the ksz_common.c and used the dev_ops for
register read and write. This function can also used for the lan937x
part. Hence making it generic for all the parts.
Signed-off-by: Arun Ramadoss <arun.ramadoss@microchip.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Link: https://lore.kernel.org/r/20220424112831.11504-1-arun.ramadoss@microchip.comSigned-off-by: Paolo Abeni <pabeni@redhat.com>

de6dd626

net: phy: LAN937x: add interrupt support for link detection · fb0a43f5

Arun Ramadoss authored Apr 23, 2022

Added the config_intr and handle_interrupt for the LAN937x phy which is
same as the LAN87xx phy.
Signed-off-by: Arun Ramadoss <arun.ramadoss@microchip.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Link: https://lore.kernel.org/r/20220423154727.29052-1-arun.ramadoss@microchip.comSigned-off-by: Paolo Abeni <pabeni@redhat.com>

fb0a43f5

25 Apr, 2022 19 commits

wwan_hwsim: Avoid flush_scheduled_work() usage · cc271ab8

Tetsuo Handa authored Apr 22, 2022

Flushing system-wide workqueues is dangerous and will be forbidden.
Replace system_wq with local wwan_wq.

While we are at it, make err_clean_devs: label of wwan_hwsim_init()
behave like wwan_hwsim_exit(), for it is theoretically possible to call
wwan_hwsim_debugfs_devcreate_write()/wwan_hwsim_debugfs_devdestroy_write()
by the moment wwan_hwsim_init_devs() returns.

Link: https://lkml.kernel.org/r/49925af7-78a8-a3dd-bce6-cfc02e1a9236@I-love.SAKURA.ne.jpSigned-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Reviewed-by: Sergey Ryazanov <ryazanov.s.a@gmail.com>
Reviewed-by: Loic Poulain <loic.poulain@linaro.org>
Link: https://lore.kernel.org/r/7390d51f-60e2-3cee-5277-b819a55ceabe@I-love.SAKURA.ne.jpSigned-off-by: Jakub Kicinski <kuba@kernel.org>

cc271ab8

net: dsa: remove unused headers · df1cc211

Marcin Wojtas authored Apr 25, 2022

Reduce a number of included headers to a necessary minimum.
Signed-off-by: Marcin Wojtas <mw@semihalf.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

df1cc211

arp: fix unused variable warnning when CONFIG_PROC_FS=n · b0e653b2

Yajun Deng authored Apr 22, 2022

net/ipv4/arp.c:1412:36: warning: unused variable 'arp_seq_ops' [-Wunused-const-variable]

Add #ifdef CONFIG_PROC_FS for 'arp_seq_ops'.

Fixes: e968b1b3 ("arp: Remove #ifdef CONFIG_PROC_FS")
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Yajun Deng <yajun.deng@linux.dev>
Signed-off-by: David S. Miller <davem@davemloft.net>

b0e653b2

net: ipa: compute proper aggregation limit · c5794097

Alex Elder authored Apr 21, 2022

The aggregation byte limit for an endpoint is currently computed
based on the endpoint's receive buffer size.

However, some bytes at the front of each receive buffer are reserved
on the assumption that--as with SKBs--it might be useful to insert
data (such as headers) before what lands in the buffer.

The aggregation byte limit currently doesn't take into account that
reserved space, and as a result, aggregation could require space
past that which is available in the buffer.

Fix this by reducing the size used to compute the aggregation byte
limit by the NET_SKB_PAD offset reserved for each receive buffer.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

c5794097

net: ethernet: mtk_eth_soc: add check for allocation failure · a00e41bf

Dan Carpenter authored Apr 21, 2022

Check if the kzalloc() failed.

Fixes: 804775df ("net: ethernet: mtk_eth_soc: add support for Wireless Ethernet Dispatch (WED)")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

a00e41bf

ethernet: broadcom/sb1250-mac: remove BUG_ON in sbmac_probe() · 60d78e9f

Yang Yingliang authored Apr 21, 2022

Replace the BUG_ON() with returning error code to handle
the fault more gracefully.
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

60d78e9f

net: mscc: ocelot: Remove useless code · 985e254c

Haowen Bai authored Apr 21, 2022

payload only memset but no use at all, so we drop them.
Signed-off-by: Haowen Bai <baihaowen@meizu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

985e254c

Merge branch 'mlxsw-line-card-model' · 5e927a9f

David S. Miller authored Apr 25, 2022

Ido Schimmel says:

====================
mlxsw: extend line card model by devices and info

Jiri says:

This patchset is extending the line card model by three items:
1) line card devices
2) line card info
3) line card device info

First three patches are introducing the necessary changes in devlink
core.

Then, all three extensions are implemented in mlxsw alongside with
selftest.

Examples:
$ devlink lc show pci/0000:01:00.0 lc 8
pci/0000:01:00.0:
  lc 8 state active type 16x100G
    supported_types:
      16x100G
    devices:
      device 0
      device 1
      device 2
      device 3
$ devlink lc info pci/0000:01:00.0 lc 8
pci/0000:01:00.0:
  lc 8
    versions:
        fixed:
          hw.revision 0
        running:
          ini.version 4
    devices:
      device 0
        versions:
            running:
              fw 19.2010.1310
      device 1
        versions:
            running:
              fw 19.2010.1310
      device 2
        versions:
            running:
              fw 19.2010.1310
      device 3
        versions:
            running:
              fw 19.2010.1310

Note that device FW flashing is going to be implemented in the follow-up
patchset.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

5e927a9f